Safety Situation Forecast of Mining System based on Dynamic Division of States

A forecast method of mining system safety situation is proposed based on dynamic division of states. On the basis of the number of total mine safety accidents, the boundary lines was determined according to the comparative analysis of residuals between estimated values based on seasonal index trend fitting model and actual number of total mine safety accidents, obtaining the dynamic division of transition states. Then weighed Markov forecast model under the influence of environment factors was built, by which, the safety situation of mining system was predicted. It was found that dynamic division can improve prediction accuracy, which enables the prediction model to better reflect the safety situation of mining system.